凯茜·伍德谈人工智能与科技巨头的未来走向

有人向你兜售一个人工智能未来，在那里你将变得过时或无足轻重。这种愿景是错误的。在Palantir，他们正在构建能帮助工人并释放其全部潜能的人工智能。美国工人是我们国家最伟大的力量。人工智能不应淘汰他们。

You're being sold an AI future where you're obsolete or irrelevant. That vision is wrong. At Palantir, they're building AI that helps workers and unlocks their full potential. American workers are our nation's greatest strength. AI shouldn't eliminate them.

它应该提升他们。Palantir在此讲述他们的故事。从工厂到医院，人工智能正将人们从苦差中解放，让他们做人类最擅长的事——创造、解决、建设。Palantir，让美国人变得不可替代。

It should elevate them. Palantir is here to tell their stories. From factories to hospitals, AI is freeing people from drudgery, letting them do what humans do best, create, solve, build. Palantir, making Americans irreplaceable.

我是亚历克斯·罗德里格斯。我是杰森·凯利。我们回来了

I'm Alex Rodriguez. And I'm Jason Kelly. And we're back

带着更多《交易》节目的剧集。

with more episodes of our show, The Deal.

我们带来了与体育界最伟大的商界人士和运动员的幕后对话。从塞雷娜·威廉姆斯和斯蒂芬·A·史密斯到比尔·贝利奇克和梅洛迪·霍布森。

We've got behind the scenes conversations with sports greatest business people and athletes. From Serena Williams and Stephen A. Smith to Bill Belichick and Melody Hobson.

我在看到上涨箭头时变得危险起来。

I got dangerous in seeing up arrows up arrows.

补上你错过的采访，每周四在Apple播客、Spotify、YouTube或任何你获取播客的地方收听全新剧集。

Catch up on the interviews you've missed and listen to all new episodes every Thursday on Apple Podcasts, Spotify, YouTube, or wherever you get your podcasts.

嘿，OddLots的听众们，我们要来华盛顿特区了。

Hey, OddLots listeners. We're coming to DC.

我们终于要实现了，乔。这将是我们首次在首都华盛顿特区举办现场节目，也是我们终于要真正讨论《琼斯法案》的时刻。

We're finally doing it, Joe. It's gonna be our first live show in Washington, DC, our nation's capital. It's also finally gonna be the time where we actually talk about the Jones Act.

我们早就说要制作一期关于《琼斯法案》的OddLots节目，这已经成了个我们从未兑现的经典老梗。但这次我们要以盛大的方式完成它

We've been talking about doing the Jones Act episode of Odd Lots for a long time, and it's become this recurring joke that we've never done. But we're gonna do it in grand style

嗯哼。

Mhmm.

因为我们将在华盛顿特区现场进行，而且实际上会是一场辩论。

Because we're gonna be doing it live in DC, and it's actually gonna be a debate.

没错。我们请到了运输研究所的莎拉·富恩特斯支持法案，还有卡托研究所的科林·格雷博持反对立场。看双方如何交锋会非常有趣。

Yeah. So we have Sarah Fuentes from the Transportation Institute. She's gonna be taking the pro side, and we also have Colin Grebaugh of the Cato Institute. He'll be taking the against side. It's gonna be really interesting to see how all of that shakes out.

除此之外，我们还将与电信泡沫时期的亲历者布莱尔·莱文对话，以及接替莉娜·汗的新任FTC主席安德鲁·弗格森。我们会讨论并购交易之类的话题，应该会是个非常精彩的夜晚。

In addition to that, we're gonna be speaking with Blair Levin, who was around during the telecom bubble. And we have Andrew Ferguson, the new head of the FTC, the one who's replaced Lena Khan. We're gonna be talking about mergers and acquisitions and all that stuff. So it should be a really fun night.

如果你想参加我们那晚的活动，时间是3月12日在奇迹剧院。访问bloomberg.com/oddlots，就能找到购票链接。期待在那里见到你。

If you want to come and join us for that evening, it's gonna be on March 12 at the Miracle Theater. Go to bloomberg.com/oddlots, and you can find the link to purchase tickets. We hope to see you there.

彭博音频工作室。播客。电台。新闻。

Bloomberg Audio Studios. Podcasts. Radio. News.

大家好，欢迎收听新一期的《奇思妙想》播客。我是崔西·阿莱威。

Hello, and welcome to another episode of the Odd Thoughts podcast. I'm Tracy Allaway.

我是乔·韦森塔尔。

And I'm Joe Weisenthal.

乔，我超爱我们做现场活动的时候。特别有意思。有时还有观众互动环节，这很棒。我们可以坐在台上显得很重要。这周我们就在一个非常重要的舞台上。

Joe, I love it when we do live events. It's very fun. Sometimes there's audience participation, which is great. We get to, you know, sit on stage and look important. And this week, we were on a very important stage.

我们参加了彭博投资峰会。这是我们年度最重要的旗舰会议，终于采访到了一位我渴望已久的人物。

We were at the Bloomberg Invest Conference. This is our flagship conference of the year, and we got to speak to someone who I've wanted to interview for quite a long time.

我当然也喜欢现场活动，但你知道，我也很享受戴着耳机的舒适感，不用操心自己的形象。

I like doing live events for sure, but, you know, I also love the comfort of my headphones, not having to worry about how I look.

我确实会担心嘉宾是否会出现。

I do get nervous about guests showing up.

哦，我也总是为此紧张，但他们每次都准时到场。我们做直播活动以来，从没有嘉宾缺席的情况。

Oh, I get nervous about that's all and they've always showed up. We've never done a live event where the guest didn't show up.

噢，你现在这么一说可要招来霉运了。

Oh, you're gonna jinx it now.

我知道。但确实如此。不过在3月4日周二那天，我们参加了彭博投资大会，有幸采访了凯茜·伍德。

I know. But that's right. But on Tuesday, March 4, we were down at the Bloomberg Invest Conference, and we got to interview Cathie Wood.

没错。凯茜·伍德是ARK Invest的创始人兼CEO，以投资特斯拉等科技巨头闻名。我们向她询问了关于埃隆·马斯克等问题，主要是想了解她对未来几年科技发展趋势的见解。她非常乐观，乔。

That's right. So Cathie Wood, famously the founder and CEO of ARK Invest, famously an investor in some pretty big tech names like Tesla. So we got to ask her, you know, questions about Elon Musk and things like that. And basically just hear what she thinks is coming down the pipeline in the upcoming years in terms of technology. She's very optimistic, Joe.

她确实非常乐观。她预期我们将迎来科技投资的黄金时代，美国商业的黄金时期，用她的话说，会超越里根时代。来听听我们对凯茜的采访吧。

She's very optimistic. She expects to see a golden age of tech investing, a golden age for American business, bigger than the Reagan era, as she called it. So take a listen to our interview with Kathy.

当前市场的焦点是过去几天的大范围抛售。但在此之前，我们看到AI领域出现了针对性下跌。这种担忧似乎凭空出现——突然所有人都在讨论中国AI模型深度求索，当时出现了剧烈抛售。在那天之前，你听说过深度求索吗？

So the big story in markets right now is a widespread sell off in the past couple of days. But even before then, we saw something specifically focused on AI. And the worries seemed to come out of nowhere. Suddenly, everyone was talking about DeepSeek, this Chinese AI model, and we saw a really intense sell off. Were you aware of DeepSeek before that day?

你认为这种情况对美国AI来说是个多大的问题？实际上，

And how big a problem do you think something like that is for US AI? Actually,

我们之前就注意到了DeepSeek。我记得那篇论文是12月发表的。我们的分析师仔细研究后认为这是个非常有趣的模型。让人惊讶的问题是：等等，构建这个大语言模型真的只花了600万美元吗？他们真的仅用高端工作站就完成了？

we were aware of DeepSeek, r one. I think the paper came out in December. So our analysts had pored over it and thought it was a very interesting model. I think the surprise and the question was around, wait a minute, did this take only $6,000,000 to build this large language model? And did they really do it on a high end workstation?

我们是否还需要这么多数据中心容量？随着时间推移，我们了解到他们之前做了大量预训练。显然他们拥有5万个GPU集群来辅助预训练。不过，包括Sam Altman和黄仁勋在内的顶尖技术专家都评论说，他们对该算法本身印象深刻。最让我们惊叹的是它的开源特性。

Are we going to need all this data center capacity after all? And I think with time, we've learned that they did a lot of pre training before. Apparently, they have 50,000 cluster of GPUs, which help with the pre training. Nonetheless, as some of our finest technology experts, Sam Altman and Jensen Huang included, they commented, they were terribly impressed by the algorithm itself. And I think what we were impressed by is it's open source.

任何人都可以使用。看吧，Meta Platforms曾是开源大语言模型的代表，现在我们有了DeepSeek。而且昨天又出现了一个叫功夫的新模型。

Anyone can use it. Here we go. Meta Platforms was really the open source platform, large language model. Now we have DeepSeek. And I think yesterday, another one came out, Kung Fu.

哦，我完全错过了这个。我最爱尝试新模型了，明天就去试试。今天是3月4日录制时，市场下跌引发担忧：效率提升是否意味着芯片需求减少或数据中心扩张动力减弱？

Oh, I totally missed that one. I love trying the new ones. I'll try it tomorrow. Right now, we're recording this March 4. When the market sold off, there was this concern like, oh, will the greater efficiency mean at the margins less demand for chips or less impulse to build out data centers?

虽然没有确凿证据，但科技巨头每年都有惊人的资本支出预算。这里那里有些蛛丝马迹。过去几周你观察到什么具体迹象，能说明除了DeepSeek之外，企业或超大规模运营商对AI的投入边际正在发生变化吗？

And there's no hard evidence of that, but the big tech companies, enormous capex budgets every single year, and they're throwing out these unbelievable numbers. There are hints here and there. Have you observed anything tangible, say over the last several weeks, that to you suggests that there is some change at the margin deep seek aside in how much businesses, hyperscalers, etcetera, are throwing at AI right now?

我还没有发现任何变化

I have not seen any change in the

没有证据表明存在一个‘The’

There's no evidence of a The

这些模型的力量是深远的。事实上，我们刚与一家大型LLM进行了对话。我们签了保密协议，抱歉不能透露具体是哪家供应商。但显然各国、教育系统和企业都在说：我们必须推进这件事。

power of these models is profound. And if anything, we just had a conversation with one of the largest LLM. We're under an NDA. Sorry, I can't talk about which one, providers. And it's clear that nations, educational systems, and enterprises are all saying, we have got to do this.

这在多方面都具有变革性。许多人都在讨论生产力和效率，但也有人关注深度研究——尤其是结合新型推理模型的深度研究，他们对我们取得的成果感到震惊。我完全没有看到任何放缓的迹象。

This is transformational in so many ways. Many are talking and thinking about productivity and efficiency. But others are thinking about deep research, especially with the new reasoning models and deep research is one of them and are just blown away by the results that we're getting. So I'm not seeing any slowdown at all.

你提到DeepSeek是开源的，这使它与其他模型形成显著差异。我知道你是开源的支持者。你如何将这点纳入投资分析？另外，OpenAI选择不开源是否是个错误？

So you mentioned the fact that DeepSeek is open source and that was a big differentiator between it and some other models. I know you're a big fan of open source. How do you incorporate that aspect into, I guess, an investment analysis? And also, did OpenAI make a mistake by not going open source?

自ChatGPT问世以来，我们一直在追踪闭源和开源模型。闭源模型确实曾领先，但若看性能提升曲线的斜率，开源实际上更陡峭。我热爱开源是因为它促进竞争、推动行业发展、加速进步。开源紧咬闭源的脚跟绝对是件好事。

Well, we've been tracking closed models and open source for quite ever since the ChatGPT moment. And what you'll see is the closed models have been ahead of the open source models. But if you look at the slope of the line of the performance improvement, open source is actually a steeper slope. So, the reason I love open source is it is helping along the competition, helping the movement along, helping it go faster. And I think that open source nipping at the heel of closed is a very good thing.

要知道，Chad GPT在2022年底才进入大众视野。客观地说，任何用过这些工具的人都会感到震撼。但你是否惊讶于：尽管工具如此卓越（虽然也有缺陷），我们却至今没看到明显的宏观影响？没有哪个经济领域或劳动力因此大规模裁员。

So, you know, Chad GPT came onto everybody's radar, I guess late twenty twenty two. And I think objectively, anyone who spends any time with these tools, in a sense, is just absolutely jaw dropping. Right? Are you surprised, however, that we're like this far into it and the tools are so, in some sense, extraordinary, but we can talk about some of the flaws and limitations, And yet we really haven't seen much of a macro impact from their use. We haven't seen some sector of the economy or the labor force get laid off.

我们也没看到可衡量的生产力大幅提升（尽管众所周知这很难衡量）。你对技术本身的存在与其相对温和的宏观影响之间这种反差感到意外吗？

We haven't seen some major surge in measured productivity gains, although that's infamously hard to measure. Are you surprised in any sense by the existence of the technology and what seems like sort of a modest macro impact?

所以我确实认为生产力在某种程度上得到了提升。你说得对，这确实很难衡量。我的背景是经济学。八十年代时，生产力就是一个大问号，我们当时就发现衡量方法存在缺陷，现在可能依然如此。

So I do think productivity has been boosted to some extent. And you're right. It's very difficult to measure this. My background's economics. And in the eighties, productivity was a big question mark, and we saw how flawed the measurements are and probably still are.

所以我只能告诉你，企业的采用速度如此之快，是因为他们看到了变化。也许他们没有裁员，但也不再招聘新人。这是因为这些AI工具，尤其是对工程师而言，大幅提高了他们的生产力。对吧？所以你就不需要再招下一个人了。

So I can only tell you the rate of uptake is so much is that enterprises are seeing a difference. Maybe if they're not laying people off, they're not hiring them. And that's because these AI tools are making their own, especially engineers, so much more productive. Right? So you just don't have to hire that next person.

如果你看过美国编程员工数量的数据，就会发现它像悬崖一样直线下降。我是说，断崖式下跌。不是像悬崖那样下降，而是直接从悬崖上掉下去了。不知道你是否看过那张图表。

If you've seen the number of coding employees in The United States has dropped like a cliff. I mean, off a cliff. It hasn't dropped like a cliff. It's dropped off a cliff. And I don't know if you've seen that chart.

随着工程师生产力不断提高，这种情况确实发生了。我们在思考的是：传统软件堆栈的哪一部分会受到影响？我们认为，在AI革命中，目前已经在一定程度上失去市场份额（虽然仍在增长）的是软件即服务。我们都在密切关注这些公司的收入增长动态，比如Salesforce.com的收入增长并没有回升。

So that very much has happened as engineers just become more and more productive. You know, one of the things that we're wondering is, okay, what part of the software stack will this, the traditional software stack, will this impact? And we think the AI revolution, what we see already losing share to some extent, it is still growing, but losing share is software as a service. And, you know, I think we're all paying very close attention to the revenue growth dynamics of these companies. Salesforce.com, revenue growth isn't picking up.

事实上，它还在持续放缓。我认为下个季度会从9%降到7%，这不该是正常现象。所以我想这可能就是你提到的情况。等等——

In fact, it continues to decelerate. I think it's next quarter will be, I think, 7% down from 9%. That's not what's supposed to happen here. And so I think that might be what you're referring to. Wait a minute.

这些收入增长的动态在哪里？它们从何而来？我认为有很多企业家在车库或研发中心创造着下一个重大突破。我们看到了AI带来的两个深远影响，这是我们长期研究得出的结论。

Where are the top line dynamics? Where are they coming from? Well, I think we have a lot of entrepreneurs in a lot of garages or in R and D centers who are creating the next big thing. And, you know, we look at two profound ramifications of AI that we understand. We've been researching them for so long.

地球上最大的AI项目是机器人出租车和自动驾驶网络。我们认为这将在未来五到十年内创造8到10万亿美元的全球收入，而现在还是零。这就是所谓的具身AI。我们相信，AI最深刻的应用将出现在医疗领域。测序技术、人工智能和CRISPR基因编辑等新技术的融合，已经开始治愈疾病。

The largest AI project on Earth is robotaxis, autonomous driving networks. We think that's going to drive 8 to $10,000,000,000,000 in revenue globally in the next five to ten years, up from zero now. So that's called embodied AI. The most profound application of AI, we believe, is going to be in health care. And the convergence of sequencing technologies, artificial intelligence, and new technologies like CRISPR gene editing are already curing diseases.

镰状细胞病和β地中海贫血已被治愈。CRISPR Therapeutics公司拥有这一疗法，现已开始创收。人们对此难以置信，因为这是前所未有的。但医疗保健领域的研发爆炸式增长，是我前所未见的。

Sickle cell disease and beta thalassemia cured. CRISPR Therapeutics has that cure. It is generating revenue now. People find this one very hard to believe because it hasn't happened before. But the R and D explosion in health care is like nothing I've ever seen.

我们相信我们将重返医疗保健的黄金时代。我始于八十年代，当时基因泰克崛起并创造了那时的医疗黄金时代。那时的研发回报率在30%左右。如今，对整个制药生物技术领域而言，回报率已降至4%左右。我们认为，凭借所有这些新工具以及医疗保健领域研发与发现中增加的惊人生产力，

And we think we're going to return to the golden age for health care. I started in the eighties when Genentech had taken off and created the golden age for health care back then. Returns to R and D back then were in the 30% range. Today, for the broad based pharma biotech field, the returns are down in the 4% range. We think with all of these new tools and the incredible productivity being added to research and discovery in the healthcare space.

我们将重返研发回报率可能超过40%的黄金时代。

We're going back to the golden age where returns on R and D could be 40% plus.

硅谷向你兜售的未来是你会被淘汰，或者更糟，变得千篇一律。在Palantir，他们见证着截然不同且革命性的景象——从国防工业基地的再工业化，到造船工人提速建造，再到一线工人提升生产力，AI正在全美范围内改变工作形态。AI并非取代美国工人或将他们压制成同一模板，而是释放每个人不可替代的特质：他们的判断力、技艺和创造力。

Silicon Valley is selling you a future where you're obsolete or worse identical. At Palantir they're witnessing something different and revolutionary From re industrializing the nation's defense base to shipyard workers building faster and frontline workers boosting productivity, AI is transforming work across the nation. AI is not replacing American workers or flattening them into conformity. It's unleashing what makes each one irreplaceable. Their judgment, their craft, their creativity.

当美国工人更强大地展现自我时，他们就掌握了未来。Palantir，让美国人变得不可替代。

When American workers become more powerfully themselves, they own the future. Palantir, making Americans irreplaceable.

我是亚历克斯·罗德里格斯。我是杰森·凯利。

I'm Alex Rodriguez. And I'm Jason Kelly.

我们带着更多《交易》节目回来了。

And we're back with more episodes of our show, The Deal.

我们深入幕后，与体育界最杰出的商界人士和运动员对话，从塞雷娜·威廉姆斯和斯蒂芬·A·史密斯到比尔·贝利奇克和梅洛迪·霍布森。

We've got behind the scenes conversations with sports greatest business people and athletes. From Serena Williams and Stephen A. Smith to Bill Belichick and Melody Hobson.

我在看到上涨箭头时变得很危险，上涨箭头。

I got dangerous in seeing up arrows, up arrows.

补上你错过的访谈，每周四在苹果播客、Spotify、YouTube或任何你获取播客的平台收听全新节目。

Catch up on the interviews you've missed and listen to all new episodes every Thursday on Apple Podcasts, Spotify, YouTube, or wherever you get your podcasts.

既然你提到了无人驾驶出租车，我们得聊聊特斯拉。

Since you mentioned robotaxis, we gotta talk about Tesla.

当然。

Of course.

显然，这是你投资组合中的重要部分。说得委婉些，埃隆·马斯克最近似乎非常忙碌。作为投资者，你是否会担心他可能分心，无法专注于公司的日常运营？

Obviously, a big component of your portfolio. Elon Musk seems very busy nowadays, to put it mildly. As an investor, do you worry at all that he is perhaps distracted from ostensibly the day to day running of the company?

实际上我们从一开始就不断收到这个问题。先是特斯拉，然后他又创办了所有这些其他公司，对吧？人们总在问：这难道不会让你担忧吗？我先回答这个问题，然后再讨论政府层面的影响。

We've been getting this question practically since the beginning. So Tesla and then he starts all these other companies. Right? And people are saying, does that not concern you? So I'll answer that first, then we'll bring in the government overlay.

我们对此不担心的原因是，埃隆·马斯克可能是我们这个时代的发明家，他深知我们正处于由人工智能推动的技术最深刻融合时期。他明白，在这场竞争中最终胜出的公司必须具备：第一，深厚的领域专业知识；第二，认真对待AI并持续投入；最重要的是，它们拥有别人没有的专有数据。想想所有这些公司源源不断产生的数据。

The reason it doesn't concern us is Elon Musk is probably the inventor of our age, but who understands that we're in the midst of the most profound convergence among technologies really catalyzed by AI. And he understands that the name of the game in terms of who's going to win all this through all of this are those companies that, number one, have deep domain expertise. They take AI seriously and they're investing in it. And perhaps most important, they have data that no one else has, proprietary data. Think of all the data spewing out from all of these companies.

就连Neuralink也在收集生物数据。而医疗健康领域正经历最爆炸性的数据增长。我们体内有37万亿个细胞，每个季度都会更新。现在我们有单细胞测序技术，可以结合AI来解锁生命、健康和死亡的奥秘。这正是我们要做的事情。

Even Neuralink, that's biological data. And the most prolific data explosion out there is in the health care space. We have 37,000,000,000,000 cells in our body, and they turn over every quarter. And now we have something called single cell sequencing that we can combine with AI to unlock the secrets of life, health, and death. And that's what we're going to do.

他深谙这一点。Neuralink是其中的一部分。好的。在政府部门的覆盖层面...我同意。

He understands that. Neuralink's a a part of this. Okay. Overlay overlay in the government sector. I agree.

他确实在为自己的国家做贡献。我知道他坚信这一点。

He's doing something certainly for his country. I know I know he believes that.

他确实发了很多推文。

He's tweeting a lot. That's for sure.

他什么？

He's what?

他发了很多推文。

He's tweeting a lot.

他一向发很多推文，对吧？他总是发很多推文。无论如何，通过我们风险基金持有的他的公司，包括私有的以及特斯拉，我们在季度报告与管理层的对话中发现。

He always has tweeted a lot. Right? He always has tweeted a lot. They anyway, so he's I I think what we have found with his companies and we own in our venture fund, the private ones, as well, of course, as Tesla. And as we go through the quarterly reports dialogue with management.

对我们而言，关键在于他始终聚焦于那些构成竞争壁垒的技术核心。埃隆的专长在于，一旦发现瓶颈就会亲自介入并打破它。他运用第一性原理思考，并聚集了一群渴望攻克世界最难项目的商界人士和工程师——那些能改变人类生活工作方式的重大课题。狗狗币是另一个大项目。

Critical to us is that he is keeping his eye on the technology balls that are his competitive or barrier to entry. And what Elon is expert at doing is if there's a bottleneck, he'll go in there and blow it up. And he will use first principles thinking. And he's surrounded himself by business people and engineers who want to work on the hardest projects in the world, the hardest projects that are going to help transform the way we live and work and so forth. Doge is another big project.

这并非他的全职工作，尽管外人可能不知道。我们已与相关方沟通过，虽然近期与埃隆交流不多，但其他重要决策者都保持着高效运转。当前政治局势确实影响了销量，这是事实。

It's not his full time job, even though one would not know that. But we have talked to our counterparts and, you know, we aren't talking to Elon as much these days, but to other very important decision makers. And they're really not skipping a beat. Now, the politics of what's going on have hit sales. And so, yes, that is true.

这正是原因所在，我们早有预料。今年实际上会有两三次产品更新——Model Y作为全球最畅销车型将迎来全面改款，这一进程已在全球范围内启动。

That is why. And we knew that was going to happen. So there are a couple of calls this year, actually, probably three. We knew Model Y was going to be completely refreshed, largest selling car in the world. That is beginning to happen throughout the world.

若参照Model 3的改款效果，这次应该会很成功。但更重要的是上半年将推出的低价车型，预计3万美元以内（算上补贴会更低），这将为特斯拉打开全新市场。正如埃隆所言（尽管在政治环境下人们不信），问题从来不是需求不足。

And if the Model three refreshes any indication, should be this should work out very well. Perhaps more important is the lower cost car that they are going to put out in the first half of this year. So $30,000 or less, and think less, especially with different credits. This is going to open up Tesla to a whole new market. As as Elon says, and people don't believe him in this political dynamic, but its problem isn't demand.

很多人一直期待拥有特斯拉，只是此前无力承担。他们对即将拥有首辆特斯拉感到兴奋。这是第二点。第三点，如各位所知，我们密切关注着自动驾驶进展。

There are people who have been waiting for a car. They just can't afford it. And they're very excited to have their first Tesla. So that's the second thing. The third thing, and we've watched this very closely, as you all know, autonomous.

我们确信六月份在奥斯汀的发布将是重要里程碑。分析师现在必须将自动驾驶对特斯拉的影响纳入估值模型。那些仅将特斯拉视为电动车制造商的人需要重新评估——其自动驾驶网络/平台的毛利率将达70%-90%，而当前电动车毛利率仅15%左右。这将是颠覆性的转变，特斯拉正在转型为软件服务模式。

We do believe, well, they're launching in Austin in June, another important milestone. Now analysts have to integrate into their models what autonomous will mean for Tesla. And so anyone who's been viewing Tesla as an EV manufacturer is going to have to go back to the drawing board and realize that the gross margins of its autonomous network, its autonomous platform, will be in the 70% to 90% range, whereas their EV gross margins are in the mid teens right now. That's a double take. This is turning into a software as a service model.

我认为最终这将吸引科技投资者和分析师关注这只股票。他们理解SaaS模式，以及它与电动汽车模式的区别。关于自动驾驶中AI机遇的另一特点是赢家通吃。我们坚信特斯拉在美国市场处于领先地位。任何体验过Waymo汽车的人——比如我在波斯尼亚的经历——都会明白。

And that finally, I think, will bring technology investors and analysts into this stock. They understand SaaS and how different the model is compared to an EV model. And the other thing about the AI opportunity in autonomous is it's winner take most. And we do believe that Tesla will be in the is in the pole position here in The United States. Anyone who's tried a Waymo car, as I have in Bosnia.

我们都体验过。

We all have.

Waymo的忠实粉丝。

Big Waymo fans.

Waymo的忠实粉丝。

Big Waymo fans.

铁杆粉丝。如果你看过我们年度报告《大构想2025》，就会明白原因——向Waymo致敬。确实，乘坐体验很棒。但从Waymo的经济效益来看，他们的车辆完全不划算。

Big, big fans. If you look at Big Ideas 2025, which is our annual report, you will see why and kudos to Waymo. Agree. It's a delightful ride. But in terms of the economics for Waymo, their car is uneconomic, totally uneconomic.

如果他们不打算承受巨额亏损，将很难实现规模化扩张。这一点在《大构想2025》中有详细阐述。

It's going to be very difficult for them to scale without deciding to lose a lot of money. Right. And you'll feel you'll find that delineated in Big Ideas 2025.

既然你提到股票和科技投资者回归的话题，我们来聊聊科技股——显然它们上周遭受重创。但整体来看，大选后曾有过疯狂反弹，比如十二月的几次峰值，之后持续下跌。从市场表现来看，科技股抛售背后是否存在宏观因素？

Since you mentioned the stock and the idea of, like, tech investors coming back, let's talk about tech stocks because, obviously, they've gotten really hit hard over the last week. But overall, there was that furious post election rally sometime, you know, the various peaks of in things December. There's been this decline. What's going on? Is there a macro story behind the tech sell off when you look at what how the markets behaved?

你的答案是什么？

What's your answer?

嗯，我认为恐惧与贪婪之间总存在权衡。选举后第二天，市场就开始大幅扩展，不再局限于六大巨头，而是转向我们这类股票。原因包括放松监管。监管或放松监管总体上对创新构成阻碍，甚至联邦贸易委员会不允许并购，也不允许通过战略定价发现新创新的价值，比如‘这项新创新将价值连城，我们需要它’。

Well, I think, I mean, fear and greed is a constant trade off. And I think after the election, the day after the election, the market started broadening out enormously away from just the mag six towards our kind of stock. And the reasons for that include deregulation. Deregulation or regulation has been a menace for innovation generally, but even the FTC not allowing M and A and not allowing strategic price discovery to say, hey, this new innovation is going to be worth a lot and we need it.

现在联邦贸易委员会仍沿用同样的并购指导原则。什么？哦，联邦贸易委员会维持着并购指导方针。所以我们还没看到这波并购大潮。

Now the FTC is keeping the same merger guidelines. Pardon? Oh, The FTC is maintaining the merger guidance guidelines. So we haven't seen this big M and A wave.

总之，我认为我们会看到的。感觉放松监管对这届政府的执政纲领至关重要，是最重要的变量之一。因为想想之前的情况——即便企业间不存在直接竞争，也不允许并购。他们禁止了太多并购，导致大公司可以高枕无忧，股东们也不希望公司收购任何资产，以免影响股票回购或利润分配等。

Anyway, I think I think we will see it. Think deregulation is critical to this administration's mandate, it feels. And it's one of the most important variables. Because if you think about what was going on before, no M and A, even if companies didn't compete directly with one another. They disallowed so much M and A that the big companies kind of could sit back fat, dumb and happy and their shareholders didn't want them to buy anything because that would take away from their own, whether it's share repurchases or profit sharing and so forth.

因此我认为这届政府将从监管角度为创新提供绝佳的发展跑道。背后还有层因素——正如你所想，我们之前讨论的深度求索公司代表的中国正紧追我们。特朗普政府极具竞争意识，始终紧盯中国。所以这其实是件好事。

So I think that this administration is going to provide a really beautiful runway from a regulatory point of view for innovation. And I feel that what is also behind this, as you might imagine, China with deep seek, as we were talking about, okay, they're on our tail. Right? Well, the Trump administration is extremely competitive and has China in focus, shall we say. So this is this is a very this is a good news thing.

那我们行动起来吧。市场起飞还有诸多原因，但我认为税率普遍下调是关键——作为部分关税的抵销措施。我理解关税，但不喜欢关税。关税本质就是税。

So let's let's do this. The other reason I think the market took off or there are many reasons, but I think tax rates coming down broadly, which I think they will as an offset to some of the tariffs. And I understand tariffs. I don't like tariffs. Tariffs are taxes.

但如果你听凯文·哈萨德的分析，会发现这里正在形成某种交换条件：想想我国建国初期，政府全部收入都来自关税。全部。而现在关税占比极低。他们可能在进行某种再平衡游戏。关键线索在于——最先宣布的是哪些税率的下调？

But if you listen to Kevin Hassard, it seems there's a quid pro quo developing here where, wait a minute, in the early days of our country, all of government was funded by tariffs. All of it. And now very little of it. I think they might be into a little bit of a rebalancing game. And what the clue there is, in terms of tax rate reductions, what are the first ones they've announced?

小费、社会保障和加班费。这些对中低收入群体非常有吸引力，对吧？我认为在这种动态下，全面降低税率会更受欢迎。他确实如承诺般在关照普通人。

Tips, Social Security, and overtime. Those are very appealing to the lower to middle income demographic. Right? I think lowering all tax rates are gonna be is going to be much more acceptable with that kind of dynamic at work as well. He's looking out for the little guy like he said he would.

对吧？作为阿特·拉弗的学生，我认为降低税率、放松监管将重现类似里根革命的局面，但规模会更大。因为现在有五大创新平台和14种不同技术，而当年只有个人电脑。如今我们有机器人技术、能源存储、人工智能、区块链技术和多组学测序。

Right? And so I think as a student of Art Laffer, lowering tax rates, deregulation is we think going to, is going to recreate something like the Reagan revolution, but I think it's gonna be bigger. It's gonna be bigger because there are five innovation platforms now, 14 different technologies, whereas back then it was a PC. It was a PC. Now we have five robotics, energy storage, AI, blockchain technology, multiomic sequencing.

五大领域同时发展，涉及15种不同技术，并且正在融合。自动驾驶出租车网络、机器人技术与能源存储和人工智能的融合，每个领域都有自己的S曲线，现在它们将相互促进。

Five at the same time. They involve 15 different technologies, and they're converging. Autonomous taxi networks, convergence of robotics, energy storage and AI. Those are each each one of those has its own S curve. And now they're going to be feeding one another.

我曾亲历里根革命，那真是令人振奋的时期，是主动股权管理的黄金时代。我认为这种盛况即将重现，而且规模会更大，甚至会让当年相形见绌——尽管当年已经很了不起。

I mean, I think the Reagan revolution and I was there, and it was so enjoyable. It was the heyday golden age of active equity management, and I think that's coming back. I think it's coming back big time. I think this will dwarf that. And that was pretty good.

我是亚历克斯·罗德里格斯。

I'm Alex Rodriguez. And I'm Jason Kelly.

我是杰森·凯利。

And we're back with more episodes of our show, The Deal.

欢迎回到我们的节目《交易》更多精彩内容。我们将带您直击体坛顶尖商业人士和运动员的幕后对话，从塞雷娜·威廉姆斯、斯蒂芬·A·史密斯到比尔·贝利奇克和梅洛迪·霍布森。

We've got behind the scenes conversations with sports greatest business people and athletes. From Serena Williams and Stephen a Smith to Bill Belichick and Melody Hobson.

我看到上涨箭头时变得危险起来。

I got dangerous in seeing up arrows up arrows.

补上你错过的访谈，每周四在苹果播客、Spotify、YouTube或你获取播客的任何平台收听全新节目。

Catch up on the interviews you've missed and listen to all new episodes every Thursday on Apple Podcasts, Spotify, YouTube, or wherever you get your podcasts.

我想问一个关于你投资策略的概括性问题。我知道你强调过你在长期押注变革性技术，比如我们讨论过的人工智能或机器人出租车。我的问题是，在某个时间点，这些技术的承诺必须实现并真正变现。你会给自己设定盈利期限吗？还是你心里有个大致的时间框架来预期回报？

I wanna ask a a sort of general question about your investing strategy. And I know you emphasize that you're making long term bets on transformational technology like AI, which we've been discussing, or robo taxis. I guess my question is, at some point, the promise of that world has to come to fruition and actually be monetized. Do you ever set yourself deadlines for positive returns, or is there a time frame you have in your mind for when this will pay off?

我们的投资时间跨度是五年。我们研究方式中非常关键的是，决定这些技术推广速度的最重要变量是产量单位。现在，产量单位与所谓的莱特定律有关。不知道你是否想让我详细解释，这是摩尔定律的一个相关定律。

So we our our investment time horizon is five years. What's very important about the way we do our research, the most important variable in terms of determining how quickly these technologies are going to scale is units. Now, unit and something called Wright's Law. I don't know if you want me to go into it. It's a relative of Moore's Law.

这是理解每项技术相关成本下降速度的方法。因此我们对所有这些技术的成本下降动态都有很好的把握。过去五年发生的最重大事件是什么？疫情期间产量增长暴跌。随后我们面临了所有这些严重的供应限制。

It's a way to understand how quickly the costs associated with each technology are falling. So we have had a good sense of all of these technologies' cost decline dynamics. What was one of the biggest things that happened over the last five years? Unit growth plunged during COVID. And then we faced all of these massive supply constraints.

这影响了我们部分技术的变革速度。如今我们已经度过了那个阶段。实际上，过去四年我们接连克服了三大逆风，这些对我们的策略造成了严重冲击。首先是新冠疫情带来的繁荣与萧条周期及其引发的各种过度行为。

That hurt the rate of change for some of our technologies. We're on the other side of that. We are on the other side of that. In fact, we're on the other side of three major headwinds over the last four years that really hurt our strategy. First was the boom bust associated with COVID and all of the excesses around that.

其次是利率问题——这是非常重要的应对措施。短短一年多时间里利率激增24倍，这对整个体系造成了巨大冲击。但高利率总会损害我们的策略吗？完全不是这样。

Second, interest rates, very importantly, a response. 24 fold increase in little more than a year's time. That was a major shock to the system. Now do higher interest rates always hurt our strategy? Not at all.

事实上，2017和2018年是我们表现最佳的年份之一，一个处于市场上涨期，一个在下跌期而我们逆势增长。这两年利率都在上升。我认为我们刚经历了一段非常特殊的时期。如今利率阻力已经结束——今天长期债券收益率达到4.12%，虽然不知道收盘价，但几个月前谁能预料到这个水平？

In fact, 2017 and '18, we had some of our best years, one in an up year, one in a down year for the market when we were up. Interest rates growing up both years. I think we've just been through a very unusual circumstance. So we're done with the interest rate headwind, and we're done if you think about it today, the long bond yield hit 4.12%. I don't know where it ended, but who expected that a few months ago?

这预示着某些信号，我稍后会详细分析。但利率不会继续上涨，我们坚信这一点。其次是市场向'六大科技巨头'的集中趋势，这种对大盘股、现金流充裕企业的追捧真正始于2018、2019年，当然还要沾点AI这类性感概念对吧？

That's telegraphing something, and I'll get into that in just one minute. But interest rates, they're not going up. We do not believe they're going up. Second was the concentration in the market towards the Mag-six. And that really started after 'eight, 'nine, this desire for large cap, lots of cash, and yes, touches something sexy like AI, right?

这种集中化愈演愈烈。我们从未见过比这更极端的市场集中，甚至大萧条时期都没出现过——那时企业面临的是生死存亡的二元选择。当前市场同样存在根本性恐惧，资金扎堆少数股票，说明恐慌情绪弥漫。大选的首要影响就是部分恐惧得以消散，当然现在又有了新的担忧点，这个可以稍后讨论。

So that went into overdrive. We've never seen a more concentrated market in our history, not even the Great Depression, which was a binary, will this company survive or not back then. So to see the same kind of underlying fear and crowding into a few names tells me there's been a lot of fear out there. I think the the first order impact of the election was, some of that fear can dissipate. Now we have a whole new set of fears, but we can talk about that in a second.

我们认为市场已开始并将持续走向多元化。若继续向六大巨头集中，这绝非健康市场。历史表明，极端集中后只有两种结局：要么像科技泡沫破裂、70年代初'漂亮50'终结那样进入熊市，要么转向全面牛市。

So we think that the market will, it has started and will continue to broaden out. This market, if it continued towards Mag six, not a healthy market, just not. By definition, two things happen after a major concentration. One of two things. Either a bear market like tech and telecom bust and early 70s, the end of the Nifty Fifty, or the other four major episodes of concentration.

另外四次重大集中时期最终都演变为全面牛市。我们认为这个转折已开始，过程中会有反复。最令人意外的是估值阻力消失——采用企业价值/EBITDA指标（规避财务操纵），经股权激励和研发支出调整后，我们的投资组合估值已在日元套利平仓期间达到市场平均水平。

The other four major ones ended up in bull markets that broadened out. We think that has started. There will be to and fro's. Maybe the most important and surprising to many people of the headwinds, which we are no longer facing, is valuation. If you look at enterprise value to EBITDA, which is our chosen metric, so the entire cap structure divided by EBITDA, which is not subject to financial engineering, you'll see that our portfolio and we worked with S and P adjusting for SBC and R and D, and we can go into that if you want.

经过最近几周波动，我们再次接近该水平。相对标普500，我们的投资组合估值处于历史最低位。当前市场震荡的真正原因是衰退预期——自美联储激进加息以来，我们始终认为过去三年处于滚动式衰退中。

Our portfolio basically hit a market multiple during the yen carry trade unwind. And after these last few weeks, we're getting close there again. Relative to the S and P, our portfolios really are at the low point in terms of that valuation metric throughout all of our history. So the valuation headwind is gone. I think what's shaking the market up right now is a recession.

房地产市场印证了这点：汽车业低迷，中小企业遭受重创（曾面临信贷紧缩），过去三年净利润下降30%。各行业的中小企业相继陷入困境，这种连锁反应持续发生。

Now, we have been saying since the Fed jacked rates up so quickly that we've been in a rolling recession for the last three years. And housing, the housing market certainly agrees with that. Autos, punk, small businesses have been decimated. They couldn't get credit for a time there. Their net income is down 30% over the last three So one sector after another gave way with small and medium business.

确实，就业的核心就在于此，对吧？最后的关键在于消费者。沃尔玛刚刚释放信号，表明我们正开始失去消费者。而沃尔玛此前一直表示高端消费是其业绩超预期的增长点。今天塔吉特和百思买的情况也是如此。

Really, that's the backbone of employment, right? The last shoe to drop is consumer. Walmart just telegraphed we're beginning to lose the consumer. And Walmart had been saying high end had been a source of their incremental surprises to the upside. Target and Best Buy today.

所以我认为我们正处于最后阶段。问题出在消费者身上。为什么会这样？我认为货币流通速度正在急剧放缓。实际上，如果你观察季度数据，第四季度已经出现下滑，而且看起来还会再次下降。

So I think we're at the last leg. It is the consumer. And why is this happening? I think the velocity of money is slowing down dramatically. And in fact, if you look at sequentially, it dropped in the fourth quarter, and it looks like it'll drop again.

这意味着什么？意味着人们开始捂紧钱包。为什么？我要说的是，如果把联邦、州和地方政府以及医疗教育领域的准政府机构都算上，大约30%的人都在担心：'我不知道我的工作是否安全'。

What does that mean? It means people are holding on to their money. Why? Well, about I'm going to say if you include federal, state and local government and quasi government in the health care and education space, we're probably looking at 30% of the people out there saying, I don't know if my job is safe.

没错。

Right.

此外还有高收入阶层的人群在思考：'等等，人工智能能替代我的大部分工作。这到底怎么回事？'编程岗位的减少就印证了这点。所以当前存在普遍的不确定性。但这会引发什么后果？

And then you've got another layer of people out there in the higher income end of the spectrum saying, wait a minute, AI can do a lot of my job. What's going on here? And you see that with the coding fall off. So you've got uncertainty right now. But what is this going to do?

如果我们真的出现GDP负增长，这将为特朗普政府和鲍威尔主席提供各种政策操作空间。我们已经看到长期利率下降。这说明了什么？没错，实体经济正在显现疲态。

It's going to give president Trump's administration and chairman Powell all kinds of degrees of freedom if we do have negative GDP growth. We're already seeing long rates coming down. What's that telling us? Yep. Real activity is coming in.

但我认为未来的震撼弹——考虑到消息来源，我强调这个观点很久了——是通胀率将惊人地低于预期水平。

But I think the shocker going forward, consider the source, I've been saying this for a while, is that inflation is going to surprise shockingly on the low side of expectations.

凯西·伍德，非常感谢您参加彭博投资的《OddLots》节目。

Kathy Wood, thank you so much for joining OddLots at Bloomberg Invest.

谢谢。非常感谢你们的邀请。谢谢。

Thank you. Thank you very much for inviting me. Thank you.

以上是我们与ARK Invest首席执行官兼创始人凯西·伍德的对话，录制于彭博投资大会现场。我是特雷西·阿洛韦，您可以在Tracy Alloway关注我。

That was our conversation with the CEO and founder of ARK Invest, Kathy Wood, recorded live at the Bloomberg Invest Conference. I'm Tracy Alloway. You can follow me at Tracy Alloway.

我是乔·维森塔尔，您可以在The Stalwart关注我。请关注我们的嘉宾凯西·伍德，她的账号是Kathy D Wood，并查阅ARK Invest的所有研究报告。也请关注我们的制作人：卡门·罗德里格斯（Carmen Rodriguez）、卡门·阿门·达谢尔·贝内特（Dashbot）和凯尔·布鲁克斯（Kale Brooks）。

And I'm Joe Wiesenthal. You can follow me at the Stalwart. Follow our guest, Kathy Wood. She's at Kathy D Wood, and check out all of the writing that they do at ARK Invest. Follow our producers, Carmen Rodriguez at Carmen Armen Dashel Bennett at Dashbot and Kale Brooks at Kale Brooks.

更多《OddLots》内容，请访问bloomberg.com/oddlots，我们所有节目集数和每日通讯都在那里供您订阅。若想参与讨论人工智能、特斯拉、市场等话题，欢迎加入我们的Discord社区，听众们全天候都在那里交流。链接：Discord.gg/oddbots。

For more OddLots content, go to bloomberg.com/oddlots where we have all of our episodes and a daily newsletter that you can subscribe to. And if you wanna chat about all of these topics, including AI, including Tesla, including the market, check out our Discord where listeners are hanging out twenty four seven talking about all these things. Discord.gg/oddbots.

如果您喜欢《Odd Lots》，喜欢我们的现场录制活动或亲临现场的机会，请在您喜爱的播客平台给我们好评。记住，彭博订阅用户可享受完全无广告的节目收听体验，只需在Apple Podcasts找到彭博频道并按指引操作即可。感谢收听。

And if you enjoy Odd Lots, if you like it when we record live events or when you get to attend live events, then please leave us a positive review on your favorite podcast platform. And remember, if you are a Bloomberg subscriber, you can listen to all of our episodes absolutely ad free. All you need to do is find the Bloomberg channel on Apple Podcasts and follow the instructions there. Thanks for listening.

我是亚历克斯·罗德里格斯。我是杰森·凯利。我们回来了

I'm Alex Rodriguez. And I'm Jason Kelly. And we're back

更多我们的节目《交易》的剧集。

with more episodes of our show, The Deal.

我们将带来体育界最杰出的商界人士和运动员的幕后对话，从塞雷娜·威廉姆斯和斯蒂芬·A·史密斯到比尔·贝利奇克和梅洛迪·霍布森。

We've got behind the scenes conversations with sports greatest business people and athletes. From Serena Williams and Stephen a Smith to Bill Belichick and Melody Hobson.

我在看到上升箭头时变得危险起来，上升箭头。

I got dangerous in seeing up arrows, up arrows.

补上你错过的访谈，每周四在苹果播客、Spotify、YouTube或任何你获取播客的地方收听全新剧集。

Catch up on the interviews you've missed and listen to all new episodes every Thursday on Apple Podcasts, Spotify, YouTube, or wherever you get your podcasts.